The Agent Skills Directory

[SAFE]: The skill provides instructions for using a TikTok automation toolkit through a managed MCP server. It follows standard OAuth authentication patterns for account access and explicitly defines safety boundaries for data handling and publication actions.
[SAFE]: External references to https://rube.app/mcp are documented for service configuration and do not involve unauthorized code execution or data exfiltration.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-retrieved text which is correctly identified as untrusted private data, with explicit instructions to seek user confirmation for external communications. This reduces the risk surface for indirect prompt injection.

tiktok-automation